Visual metaphors to support diagnosis of Sleep Apnea-Hypopnea Syndrome

This paper presents algorithms that assist physicians in the task of visually inspecting polysomnography recordings to diagnose Sleep Apnea-Hypopnea Syndrome. To this end, the algorithms generate visual metaphors that simplify the identification and characterization of several pathological events that occur in polysomnographic recordings: reductions and complete cessations of respiratory effort-measured both from thoracic movement and from abdominal movement- and snoring measured through a microphone which records ambient noise. These algorithms and the visual metaphors they give support to have been implemented in a desktop tool. The tool has been validated by a physician using five different polysomnographic recordings, with satisfactory results.

keywords: Sleep Apnea-Hypopnea Syndrome, Biosignal Processing, Structural Pattern Recognition, Fuzzy Sets